🔢 mathematics

Sums of three cubes over a function field

Assuming a suitable form of the Ratios Conjecture and that the characteristic is greater than 3, this paper uses a function field version of the circle method to prove that a positive proportion of elements in $\mathbb{F}_q[t]$ can be represented as a sum of three cubes of minimal degree.

Original authors: Tim Browning, Jakob Glas, Victor Y. Wang

Published 2026-07-23

📖 3 min read🧠 Deep dive

CC BY 4.0

Original authors: Tim Browning, Jakob Glas, Victor Y. Wang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a detective trying to solve a mystery about numbers. In the world of regular integers (the counting numbers like 1, 2, 3), there is a famous, stubborn puzzle: Can every number be written as the sum of three perfect cubes? For example, $1 = 1^3 + 0^3 + 0^3$ , and $2 = 1^3 + 1^3 + 0^3$ . But some numbers, like 4 or 5, seem impossible to build this way. Mathematicians have been chasing this ghost for decades. While recent work has already proven that "most" numbers can be built from three cubes, the path is incredibly rocky and full of dead ends, relying on several heavy, unproven assumptions.

To make progress, this paper takes a detour. Instead of walking on the familiar ground of regular integers, the authors step into a parallel universe called a "function field." Think of this universe not as a place of numbers, but as a place of polynomials—expressions like $t^2 + 3t + 1$ where $t$ is a variable. In this world, the rules of arithmetic are slightly different, but the puzzles look remarkably similar. It's like practicing a difficult piano piece on a keyboard that has slightly different keys; if you can master the song here, you might learn how to play it on the real thing. The authors are using a powerful mathematical tool called the "circle method," which is like a high-tech metal detector that sweeps through this polynomial universe to find hidden treasures (solutions to equations). They are looking for a specific treasure: numbers that can be written as the sum of three cubes.

The main discovery in this paper is that, in this polynomial universe, a positive chunk of the numbers can indeed be built as the sum of three cubes, provided the "characteristic" of the field is greater than 3. This is a significant step forward because, unlike the integer case which requires a heavy mix of unproven automorphy conjectures and the Riemann Hypothesis, this result removes many of those hypotheses, relying instead on just one key assumption. The authors have to make a big assumption to get this result. They rely on a "Ratios Conjecture," which is a sophisticated guess about how certain mathematical patterns (called L-functions) behave. It's like saying, "If we assume the weather forecast is correct, then we can predict that the picnic will happen." The paper proves that if this conjecture is true, then the answer to the puzzle is "yes" for a lot of these polynomial numbers.

The authors also show that if the characteristic of the field is 3, the whole game changes: the set of numbers you can build becomes so sparse that it effectively disappears. They also rule out the idea that you can always find solutions with the "smallest possible" degree (a measure of the size of the polynomial) without making that big assumption. While they haven't solved the original puzzle for regular integers yet, they have built a very strong bridge in this parallel universe, showing that the "sum of three cubes" mystery is solvable there, provided we accept a few well-motivated guesses about the underlying patterns of the universe.

Problem Statement
The paper investigates the representation of elements in the polynomial ring $\mathcal{O} = \mathbb{F}_q[t]$ as sums of three cubes. While a folklore conjecture in number theory posits that any integer $k \not\equiv \pm 4 \pmod 9$ is a sum of three integer cubes, the situation over function fields differs significantly. For $\text{char}(\mathbb{F}_q) \neq 3$ and $q \notin \{2, 4, 7, 13, 16\}$ , Serre and Vaserstein have explicitly constructed linear solutions $x, y, z$ to $x^3 + y^3 + z^3 = k$ , though these solutions have degrees $\geq \deg k$ , whereas one might hope for solutions of degree $\sim \frac{1}{3} \deg k$ . The authors aim to prove that a positive proportion of elements in $\mathbb{F}_q[t]$ are representable as sums of three cubes of minimal degree (specifically, with $\max\{\deg x, \deg y, \deg z\} \leq \lceil \deg k / 3 \rceil$ ), assuming $\text{char}(\mathbb{F}_q) > 3$ .

Methodology
The authors employ a function field version of the Hardy-Littlewood circle method. The central object of study is the counting function $N(P)$ , which counts solutions to $x_1^3 + \dots + x_6^3 = 0$ with bounded degree. The analysis relies on the following components:

Delta Method and Poisson Summation: The counting problem is transformed into an integral involving exponential sums $S_r(c)$ over the ring $\mathcal{O}$ .
L-functions and the Ratios Conjecture: The exponential sums are linked to Hasse-Weil $L$ -functions $L(s, c)$ associated with smooth hyperplane sections of the cubic hypersurface. The core of the argument requires asymptotic formulas for the mean values of ratios of these $L$ -functions, specifically $\sum 1/(L(s_1, c)L(s_2, c))$ .
Conditional Assumptions: The proof is conditional on the Ratios Conjecture (Conjecture 3.6) for the family of $L$ -functions $L(s, c)$ . This conjecture, originally formulated for automorphic $L$ -functions over $\mathbb{Q}$ , is adapted here to the function field setting. The authors note that for quadratic Dirichlet $L$ -functions, this conjecture is known for large fixed $q$ via homological stability results (Bergström et al., Miller et al.).
Geometric Sieve and Exponential Sum Estimates: The authors develop a geometric sieve for function fields (Theorem 7.2) to handle square-free moduli and refine estimates for "bad" exponential sums (those where the modulus shares factors with the dual form $F^*$ ). They also analyze the contribution from the dual variety, identifying a bias in exponential sums when the vector $c$ lies in the orthogonal complement of a linear space contained in the cubic hypersurface.

Key Contributions and Results

Theorem 1.1: Assuming $\text{char}(\mathbb{F}_q) > 3$ $char (F_{q}) > 3$ and the Ratios Conjecture for $L(s, c)$ $L (s, c)$ , the paper proves that the following sets have positive lower density in $\mathcal{O}$ $O$ :
1. The set of sums of three monic cubes $\{x^3 + y^3 + z^3 : x, y, z \in \mathcal{O} \text{ monic}\}$ .
2. The set of "strict sums of three cubes" $\{k : x^3 + y^3 + z^3 = k \text{ is soluble with } \max\{\deg x, \deg y, \deg z\} \leq \lceil \deg k / 3 \rceil\}$ .
Removal of $\epsilon$ from Bounds: Previous work by Glas and Hochfilzer established an upper bound $N(P) = O_\epsilon(|P|^{3+\epsilon})$ unconditionally. This paper removes the $\epsilon$ in the exponent (proving $N(P) \ll |P|^3$ ) at the cost of assuming the Ratios Conjecture.
Refinement of Circle Method: The authors introduce specific refinements to the circle method over function fields:
- They restrict weight functions to avoid character twists, simplifying the Ratios Conjecture requirements.
- They utilize Poincaré duality to produce poles of exterior square $L$ -functions, rendering certain automorphy hypotheses unconditional.
- They extend bias arguments in exponential sums from diagonal forms to arbitrary non-singular senary cubic forms, utilizing geometric insights on quadric bundles (Beauville).
Theorem 11.1: The paper establishes a chain of implications showing that the Ratios Conjecture (for shifted second moments) implies the bound $N(P) \ll |P|^3$ , which in turn implies that the set of representable elements has positive lower density.

Significance and Claims
The paper claims to be the first to prove that a positive proportion of elements in $\mathbb{F}_q[t]$ are representable as sums of three cubes of minimal degree, under the assumption of the Ratios Conjecture.

Contrast with Integers: Unlike the integer case, where the problem remains open for a positive proportion of integers without strong conjectures (GRH, automorphy, Ratios), this work demonstrates that in the function field setting, the problem can be reduced to a single conjecture (Ratios) that is plausibly resolvable via homological stability for large $q$ .
Conditional Nature: The authors are explicit that the result is conditional. They state that the Ratios Conjecture is the primary hurdle. They suggest that the "homological stability framework" of Bergström et al. and Miller et al. could eventually resolve the specific form of the Ratios Conjecture needed for sufficiently large $q$ .
Limitations: The result assumes $\text{char}(\mathbb{F}_q) > 3$ . The case $\text{char}(\mathbb{F}_q) = 3$ is trivial (density 0). The case $\text{char}(\mathbb{F}_q) = 2$ is noted as requiring modifications to several lemmas and propositions, which are left for future work.
Future Work: The authors mention plans to extend these results to show a density of 1 (rather than just positive density) and to handle the characteristic 2 case. They also note that the current work focuses on keeping the paper "clean" by deferring the proof of density 1 and the full resolution of the Ratios Conjecture to future publications.

In summary, the paper provides a rigorous conditional proof that the "strict" Waring problem for cubes over function fields has a positive density of solutions, bridging the gap between unconditional upper bounds and the expected asymptotic behavior via the Ratios Conjecture.

More like this